I have not planned to write an action, I thought more of bash/python to
pull the artifacts and use existing official action for publishing, but
yeah - good idea - I might package that into reusable action that we could
use for other projects. Might be generalisable.



On Thu, Jun 20, 2024 at 7:52 PM Greg Stein <gst...@gmail.com> wrote:

> Hey Jarek ... note that we have an infrastructure-actions repository for
> "official ASF" GH Actions. If you agree with that approach, then you can
> dev/test there or we can move your tested Action there when you're ready to
> share it with others.
>
> Cheers,
> Greg
> InfraAdmin, ASF
>
>
> On Thu, Jun 20, 2024 at 7:10 AM Jarek Potiuk <ja...@potiuk.com> wrote:
>
> > Unless I hear otherwise, I **assume** there are no big reasons against
> > this. My plan is that I will add a Github Action (manually triggered,
> > limited to release managers only) which will NOT build the packages, but
> it
> > will download them from `downloads.apache.org` (or dist.apache.org for
> RC
> > packages) and publish them to PyPI. This should be really "safe" and will
> > remove the needs for us to keep local pypi keys to upload the packages.
> >
> > This will require repo reconfiguration, so I will have to - likely -
> open a
> > JIRA ticket to INFRA - once I do it, I will be happy to describe the
> steps
> > for all other projects that upload packages to PyPI and use GitHub.
> >
> > Does that make sense?
> >
> > J.
> >
> >
> > On Fri, Jun 14, 2024 at 12:14 PM Jarek Potiuk <ja...@potiuk.com> wrote:
> >
> > >
> > >> My only question is what do the users see in terms of the verified
> > >> identity that performed the release. Does it still appear to have come
> > >> from the individual maintainer? The ASF? Somewhere else? I'd only be
> > >> concerned if the answer was "somewhere else".
> > >>
> > >
> > > Currently users do not see anything. There was a discussion on Python's
> > > discord about exposing Trusted Published information in PyPI
> > >
> >
> https://discuss.python.org/t/pre-pep-exposing-trusted-publisher-provenance-on-pypi/42337
> > > as a "pre-PEP discussion". This resulted in Draft PEP 740 -
> > >
> >
> https://discuss.python.org/t/pep-740-index-support-for-digital-attestations/44498
> > > - where you will be able to upload multiple attestations when you
> publish
> > > your packages. So the thinking is that you can have multiple
> attestations
> > > of provenance of your package when you upload it to PyPI and a trusted
> > > publisher will be just one of them. So in our case we could also add
> our
> > > own signatures when we publish., This is still draft and we will have a
> > > chance of influencing the direction, I am sure. Generally Michael and
> the
> > > whole security team are on the spree of onboarding more and more
> projects
> > > to use trusted publishers and they are planning to discuss and
> > implemented
> > > more security/provenance features when they reach critical mass (from
> the
> > > discussions I had - I believe they are doing very well there - and
> > having a
> > > stories where prominent projects are on-board is going to help with
> that
> > as
> > > well.
> > >
> > > J.
> > >
> > >
> > >
> > >
> > >> Mark
> > >>
> > >> ---------------------------------------------------------------------
> > >> To unsubscribe, e-mail:
> > security-discuss-unsubscr...@community.apache.org
> > >> For additional commands, e-mail:
> > >> security-discuss-h...@community.apache.org
> > >>
> > >>
> >
>

Reply via email to